688 
ME. HOPKINS ON THE THEOET OE THE MOTION OE GLACIEES. 
Hence, then, considering the combined operation of conduction and infiltration, it 
appears that, to the depth of perhaps 30 feet, the interior temperature of a glacier 
will be 32° (Fahr.) during the summer portion of the year, but will be rather lower than 
32° (Fahr.) during the winter. For all deeper parts of the glacier it will be invariably 
equal to 32° (Fahr.). 
These results are in exact accordance with the careful observations made by M. Agassiz, 
which have already been partially referred to. Besides the winter observations above 
mentioned, he also observed the temperatures in the month of July, at depths of from 
3 to 5 metres, at 30, and at 60 metres. These temperatures were all exactly 32° (Fahr.) 
during the fortnight they were observed, with the exception of one or two very small 
and manifestly accidental variations in the more superficial observations. 
11. It follows from the preceding articles that the temperature at the lower surface 
must always be the freezing-temperature, ^. e. the ice there must be in that state in 
which the mutual cohesion of its constituent particles is less than in any other state. It 
does not follow that the glacier would not slide if the temperature of its lower surface 
were less than 32° (Fahr.) ; but that temperature is the most favourable for the motion 
of the glacier, because the most favourable to the disintegration of its lower surface, and 
the immediate conversion of the ice which forms it into water. It should be remem- 
bered, too, that it was one of the results of my experiment, that, coeteris jparilus^ the 
motion was increased by increasing the weight of the mass; e., the cohesive power of 
the ice at the bottom of the glacier will be the more rapidly overcome by increasing 
the depth of the mass, the area of its base being unchanged. Consequently the tendency 
of a glacier to descend down its bed would be indefinitely greater, cmteris paribus ^ than 
that of our experimental lump of ice down its plane. It is this enormously increased 
tendency that enables the mass of a glacier to overcome the resistance arising from the 
inequalities of the sides and bottom of its valley. We shall explain in the sequel the 
prodigious force which may thus be exerted, and the corresponding internal tensions 
Avhich would thus be produced by it. These tensions overcome the cohesion of the 
mass, the ice breaks, and the glacier obtains more freedom of motion than it could have 
in its state of greater compactness and continuity. The tendency to the sliding motion we 
are considering will manifestly be greater in the axial than in the marginal parts of the 
mass. It is there, especially, that the depth must be the greatest, and the distance from 
opposing lateral objects is likewise greatest; and, it may be added, the subglacial cur- 
rents, by which the sliding will undoubtedly be more or less facilitated, will be generally 
greatest along the central parts of the valley. 
Hence, then, it follows that, so far as the motion of a glacier depends on the sliding we 
have been considering, the velocity of its axial portions will generally be considerably 
greater than that of its marginal portions. This constitutes the most distinctive and 
important character of the observed motion of a glacier. 
Still, though the sliding motion was perfectly consistent with this observed general 
character of glacial motion, it was not sufficient to account for several striking pheno- 
